INDEX

Explanations

names of locations, possibly related to politics or diplomacy

oai_token-act-pair · gpt-3.5-turbo

specific names or terms related to various contexts or entities

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

schild

-0.73

recomm

-0.65

CLASS

-0.64

 impulse

-0.60

tip

-0.60

 caution

-0.59

 Cheong

-0.57

sow

-0.57

 Brah

-0.56

 scrap

-0.55

POSITIVE LOGITS

pillar

0.98

ulhu

0.88

emporary

0.80

arel

0.79

aign

0.78

berus

0.78

arette

0.77

aminer

0.77

ãĤ¼ãĤ¦ãĤ¹

0.77

rera

0.75

Activations Density 0.100%

names of locations, possibly related to politics or diplomacy

specific names or terms related to various contexts or entities

No Comments

No Known Activations